AI Paper Atlas

Research map / Vision-language and multimodal / Robot action and navigation

SCOUT: Planning under Occlusion via Object-Centric World Model Rollouts

Amir Hossain Raj, Xuesu Xiao

NeurIPS 2026 · Vision-language and multimodal › Robot action and navigation · approach: Memory and subgoal-based manipulation · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes SCOUT, an occlusion-aware robot planning framework that combines high-level deliberation with object-centric world model rollouts to plan manipulation under partial views in cluttered scenes.

Problem

Robots manipulating in cluttered spaces must plan from partial views where foreground objects occlude the target and free space; reactive policies miss hidden constraints while exhaustive search is computationally prohibitive in dense clutter.

Approach

Bridges high-level deliberation with grounded physical rollouts from an object-centric world model to reason about occluded constraints without exhaustive search.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use multi-view instead of single-view input (ablation)SCOUT · simulated shelfavg success (%): 69.7 -> 80.8Improved
add relational projection of planning state (ablation)SCOUT · simulated shelfavg success (%): 68.1 -> 80.8Improved
add MCTS rollout search instead of greedy depth-1 (ablation)SCOUT · simulated shelfavg success (%): 70.6 -> 80.8Improved
add LLM proposals to geometric proposals (ablation)SCOUT · simulated shelfavg success (%): 74.6 -> 80.8Improved
increase MCTS search depth from 2 to 3 or 4 (ablation)SCOUT · simulated shelfavg success (%): 80.8 -> 81.5No clear effect
increase search depth from 1 to 2 (ablation)SCOUT · simulated shelfavg success (%): 70.6 -> 80.8Improved
replace GHOST initial state with oracle geometry (ablation)SCOUT · simulated shelfavg success (%): 80.8 -> 87.8Improved
re-observe after every primitiveSCOUT · simulated shelfhidden-scene failures (%): 7.6 -> 3.3Improved
re-observe after every primitiveSCOUT · simulated shelfepisode time (s): 10.4 -> 24.2Hurt
multi-view instead of single-view GHOST reconstruction (ablation)GHOST · held-out reconstructionhidden-object recall (%): 58.9 -> 72.6Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy